BSOD, Machine Check Processor Context Corrupt, THE WORKS!

[Merlin7777]

[SOLVED] BSOD, Machine Check Processor Context Corrupt, THE WORKS!

Okay, I need some help from my follow Overclockers.

To be clear, my current rig is at STOCK CLOCKS right now.

I just built this over the weekend and installed Windows yesterday.

The Rig:
ASRock Z77 Extreme 4
Core i7 3770k
4 X 4GB G.SKill DDR3 1600
OCZ Vertex 3 SSD
1 x 500GB Seagate, 1 x 1TB WD
1 x MSI GTX 650 Ti Hawk
Asus Xonar D2X
FSP Group 700W PSU (http://www.newegg.com/Product/Product.aspx?Item=N82E16817104015)

As I said, I installed Windows yesterday. I had some trouble where the SSD locked up and windows wouldn't install. I unplugged it for an hour, reset my CMOS, and I was able to install successfully.

However, today I was beginning to install my drivers and I started receiving BSODs. The first one happened right after my video drivers finished downloading, and the others happened while my driver install disk was running (several times).

I wasn't sure if it was bad drivers in Windows or what, so I attempted to boot into Ubuntu 12.04 LTS 64bit via the Live CD. Instead of the desktop that I expected, I received this:


This is peculiar, as I was able to boot into the live CD yesterday. I even ran Memtest86+ 4.20 via USB on the RAM yesterday, no problems were detected.

Almost all of the equipment is new except for the PSU, the 500GB and 1TB hard drive, and the Asus Xonar D2X.

Lend me your knowledge!

 

[Merlin7777]

I ran some of the error codes through "mcelog --acii" on another machine and this is what I got:

CPU 5: Machine Check Exception: 5 Bank 6: be20000000011166
TSC 14394cd8bd ADDR 414898880 MISC 3240004086
HARDWARE ERROR. This is *NOT* a software problem!
Please contact your hardware vendor
CPU 5 BANK 6 TSC 14394cd8bd [at 1600 Mhz 0 days 0:0:54 uptime (unreliable)]
RIP !INEXACT! 10:ffffffff8103004d MISC 3240004086 ADDR 414898880
MCG status:RIPV MCIP
MCi status:
Uncorrected error
Error enabled
MCi_MISC register valid
MCi_ADDR register valid
Processor context corrupt
MCA: corrected filtering (some unreported errors in same region)
Data CACHE Level-2 Prefetch Error
STATUS be20000000011166 MCGSTATUS 5
RIP: mwait_idle_with_hints+0x5d/0x70}
Processor 0:306a9 TIME 1340682481 SOCKET 0 APIC 3 microcode 12

 

[wingman99]

Did you try running the motherboard out of the case on the box it came in?

 

[Merlin7777]

No I have not. But come to think of it, I was running it with the case cover open yesterday and it seemed to be okay then.

My other thought is that my FSP 700w is not providing enough amps on the 12v rails. According to this, it's only 15A per rail (http://www.newegg.com/Product/Product.aspx?Item=N82E16817104015) looking at the picture of the side label.

Is there a way I can find out how many amps are necessary for my 560ti?

 

[wingman99]

560ti Requirements
Minimum of a 500 power supply.
(Minimum recommended power supply with +12 Volt current rating of 30 Amps.)you have +12 60 Amps

 

[Merlin7777]

Oh really? I wasn't aware that PSU's are rated for all summed amps on the +12v. I thought it was just one, what with the FSP 700w picture saying that 12V2 is for PCI-E2/CPU2 and 12V4 is for PCI-E1.

Hmm, since the board I'm using is 4 + 4 ATX12v and the 12V2 shares between PCI-E2 and CPU-2, is the PSU in fact sharing 15A with the CPU AND the GPU?

 

[xrror]

That seems a bit more severe than just overheating.

BIOS update maybe?
http://www.asrock.com/mb/Intel/Z77 Extreme4/?cat=Download&o=BIOS

And even a ti560 shouldn't need more than 500w in text mode, surely?

I'm at a loss though, but something in your rig is definitely not happy (duh, I know)

Processor Exceptions. Maybe try re-seating the processor and memory? I've seen rare cases where a heatsink mount (or installation) can put weird pressure on the processor, and thus the cpu doesn't sit in the socket right.

Other strangeness is maybe the board is flexed, and you have the misfortune of cold solder joints on the board - but that's rare (you'd see this if the board only works after it "heats up" ... which is a chicken/egg problem).

I'm probably totally wrong, but the errors you're seeing scream "hardware not happy"

edit add: Just realized you said the SSD locked up. Need more detail on that, if you can describe more on how it failed might give others more to work with.

edit add again: I don't know linux very well, but those error messages seem to be the kernel flagging the cores as unstable.

"As I said, I installed Windows yesterday. I had some trouble where the SSD locked up and windows wouldn't install."

Hopefully others that are more familiar with intels 6* and 7* chipsets can shed some insight on this. I would re-seat the processor, as it acts like the processor isn't communicating correctly with the "southbridge" ... but since most of that stuff is integrated into the processor now, I'm out of my league on that.

I'd start google-foo'ing hard on those error codes though, usually linux is pretty specific (and die-hard since it even ran enough to give those codes).

 

[Merlin7777]

I updated it to the lastest BIOS (1.70) yesterday.

I have ton's of air flow throughout the case, and the BIOS were saying ~33 deg. C for the processor.

The SSD failed while installing Windows. It basically hung on the "Expanding Windows files" part, stuck at 0%. A quick inspection showed that the SSD had both it's green and red lights on, vs the normal green only. After removing all connections from the SSD and resetting the CMOS, I was able to boot up with the SSD showing all green. My BIOS recognized the drive after rebooting, and I was able to successfully install windows again.

 

[RJARRRPCGP]

Those errors are more likely an unstable core OC. More likely means more Vcore required or a core wall.

 

[Merlin7777]

To be clear, my current rig is at STOCK CLOCKS right now.

 

[wingman99]

Can you try putting your OS on a different drive? Also your PSU is ok.

 

[Merlin7777]

Okay I've tried almost everything (almost) and still to no avail.

1. I returned home and attempted to boot into windows. It succeeded, and I was able to download Prime95. I started it up and waited about 1 min. BSOD.
2. Rebooted and checked the processor temp in the BIOS, as well as the Vcore :

3. Attempted to Boot into the Ubuntu Live CD, that failed like above.
4. Removed video card, attempt Live CD boot, fail.
5. Remove Xonar D2X, fail.
6. Remove all SATA connections for hard drives (leaving the Bluray/DVD drive), still fails to boot the live CD with a kernel panic.
7. Reseated all of the RAM. Took 'em out and flipped each stick to it's opposite slot. Still fails.
8. Removed Heatsink and cleaned Arctic Silver 5 from heatsink and CPU.
All pins in the connector seem to be fine:


Overall the CPU seems fine, but I noticed this strange...substance.


Can I clean this off safely with 70% rubbing alcohol and cotton balls? I wasn't sure, so I left it on there. It's not super apparent, such as exhibit B:


Reinstalled CPU (had to reseat the heatsink, which is why some of the thermal paste seems smushed):


9. Cleared the BIOS and checked temps and vCore again:


10. Attempted to reboot into the Live CD again. It hung the first time, giving a black screen, but pushing the reset button and waiting again got me back to the kernel panic screen again.


So at this point, I figure it's either A) The mobo B) the CPU or C) the PSU. I have another PSU on the way soon (550W from johnnyGuru) so I'll try that next. I suspect it's the Mobo or that strange goop.

 

[Merlin7777]

I got the CPU from I.M.O.G, who was running it subzero for a few days. Is that where the goop (glue?) could have come from?

I'm going to try pulling out all of my ram sticks and just using one to see if that makes a difference. At this point, since nobody seems to think it's the PSU, I figure it's either the Mobo or the CPU. I really hope it's the Mobo as I don't want to have to buy another CPU.

 

[Merlin7777]

Okay, it's definitely not the RAM. Pulled all 4 sticks out, tested each stick individually, in different slots even, and still boot failures.

I moved the Bluray drive to a different SATA port that runs off a completely separate controller, and boot still fails.

Since the Mobo is posting just fine and I'm not seeing any errors when I'm in the BIOS, I'm inclined to think that it's the CPU.

Should I try getting that goop off using 70% Alcohol and some cotton balls?

 

[wingman99]

I would clean the CPU with alcohol.

 

[Merlin7777]

Yeah, I think it's my only option at this point.... If it breaks something, oh well, since it is pretty much hosed right now anyways.

 

[wingman99]

I think you have a bad CPU it did the same thing as my old 2500k BSOD.

 

[Merlin7777]

Well I.M.O.G had it before me and apparently it worked like a charm. Again, he ran it subzero, but who knows, maybe it got bumped too hard in the shipping.

 

[wingman99]

I guess all you can do is clean the CPU up and see if it solves the problem.

 

[RollingThunder]

Merlin, that stuff has to be conductive and shorting out between your pins. I've never seen those error messages before.